Right now, half of the internet's

screaming that AI is the biggest bubble

in history. The other half screaming

that it's the biggest breakthrough since

sliced bread. And honestly, I avoided

making this video for a long time,

mostly because AI is the only thing

anyone talks about anymore. But the

longer this debate spiraled, the clearer

it got to me that both sides were

arguing about the wrong thing entirely.

It's reached a point where I felt like I

couldn't just sit back and watch

anymore. So here we are. Because the

uncomfortable truth is that it doesn't

matter which half of the internet is

right. Both of them are watching the

wrong thing entirely. They're missing

what's hiding in plain sight. Everyone

in this fight is arguing about the same

thing. User growth, adoption, revenue,

how many people are using this stuff.

They're all just arguing about demand.

But demand is also the side of the coin

that can lie to you. It's noisy, it's

narrative driven, and it's a story that

changes depending on who's telling it

and what they're trying to sell you. But

every coin has two sides, and the other

side, the one nobody's talking about, is

the side that paints a much better

picture of what's actually unfolding.

It's what the world's best investors

have quietly used to track financial

bubbles for more than 200 years. And it

starts by looking at the other side of

the coin, supply. Because demand can be

faked, and we'll get more into exactly

how that happens a bit later. But you

can't fake supply. Supply takes real

capital committed up front, and it tells

the real story of what's happening. This

year alone, the biggest tech companies

in the world are on track to spend

around 725 billion dollars on artificial

intelligence. And when most people hear

a number like that, they usually just

scroll right past it without ever

stopping to picture how much money that

actually is. Because 725 billion dollars

isn't just a big number. It's an almost

incomprehensible amount of money. To put

it into perspective, if you spent 1

million dollars every single day

starting from the day Jesus was born,

all the way through the fall of the

Roman Empire, the Middle Ages, the

Renaissance, both World Wars, and all

the way up to today, you still wouldn't

have spent as much as Big Tech is

projected to spend in AI in just this

year. It breaks down to roughly $2

billion a day, $83 million an hour, or

about $23,000 every second. In just the

time you've spent watching this video,

Big Tech has already spent more than $3

million on artificial intelligence.

We've never seen a bubble of this size

before, but as Mark Twain put it,

history doesn't repeat itself, but it

often rhymes. And what we're seeing

right now is very familiar. The real

question isn't whether AI is going to

change the world That part doesn't

actually really matter. Financial

bubbles aren't required to be built on

bad technology. In fact, most of the

time, it's the opposite. ResearchGate

looked at 175 years of major

breakthroughs from 1825 to 2000 that

included innovations like the

television, the automobile, and the

internet. And they found that 37 of the

51 major breakthroughs in technology

triggered a massive speculative

financial bubble. As billionaire

investor George Soros said, "Stock

market bubbles don't grow out of thin

air. They have a solid basis in reality,

but reality is distorted by a

misconception." If you look throughout

history, it's the same pattern repeating

itself over and over, and it always ends

the same way, which is why the best

investors have stopped watching the

demand side and started watching the

supply side. Because when you follow the

paper trail, you get to the truth. It

starts with a framework that is rooted

in some of the oldest ideas in economics

going all the way back to Adam Smith in

the 1700s. "The increase of stock, which

raises wages, tends to lower profit.

When the stocks of many rich merchants

are turned into the same trade, their

mutual competition naturally tends to

lower its profit. And when there is a

like increase of stock in all the

different trades carried on in the same

society, the same competition must

produce the same effect in them all."

When Adam Smith says stock, he is

referring to capital and productive

assets. So, translate this out of 1700s

powdered wig talk, and what Adam Smith

is really saying is that money naturally

flows into industries where there are

perceived high returns, which means more

competition enters, which ultimately

drives returns down for everyone in the

industry. And this idea has been

modernized and expanded upon into a

framework with a new name, the capital

cycle theory. It ignores the noise

involved with demand and just focuses on

what actually matters, the capital

flowing through the supply side of an

industry. And after I explain how it

works, you'll start to notice the same

patterns repeating across stock market

bubbles throughout history. The capital

cycle moves through an industry in four

steps. First, the flood. A new

technology shows up, the returns look

sexy, and capital comes pouring in to

chase them. Step two is the boom. Money

keeps pouring in, competitors pile in

faster than the demand can grow, and the

more everyone builds, the more the

returns start to slip. Step three, the

collapse. All that new supply finally

crushes the very returns that attracted

it in the first place. Prices fall,

companies fail, the industry

consolidates, and a lot of people lose a

lot of money. And finally, the

inheritance. This is where the real

money gets made. Capital flees the

industry, the builders go bankrupt, and

somebody walks into the wreckage, buys

the assets for pennies on the dollar,

and collects the value the original

builders in the industry paid for.

That's the capital cycle theory. The

idea is simple, returns are driven by

supply, not demand. The best

opportunities show up where capital is

fleeing, the worst where it's flooding

in. And if you go back through history,

you'll see the capital cycle showing up

again and again. But before I show you,

here's Oaktree Capital co-founder and

billionaire investor Howard Marks

breaking down how history repeats itself

in the markets.

>> Most people ignore history.

Most booms

and I you know, I've I've probably lived

through about a half a dozen real booms

uh in my 50 years in the business um

are usually about something new.

And

the internet in '99, the Nifty 50, Xerox

in '69, whatever it might be, subprime

mortgage securities.

>> [snorts]

>> And

the people who get excited about it, who

who who cotton to it, who uh

are uh intoxicated by the positives and

willing to ignore the negatives,

um

if you if you say to them, you know,

well, that kind of that happened 20 and

40 years ago and and it it ended badly,

what they say is use the four worst

words in the world.

It's different this time.

Uh the rules of the past don't apply.

The You know, if

Yes, the average PE ratio historically

has been 16, but now it's 32 and that's

okay because the internet has changed

the world.

>> So, let's rewind to one of the earliest

examples of the capital cycle theory

playing out from start to finish. Take

yourself back to the 1830s. The first

Industrial Revolution is in full swing

in Britain. Mines and factories need to

move coal and iron faster than

horse-drawn wagons allow them to. Then

new railway technology shows up and

changes everything. And where excitement

is, the money follows. This is the

flood. Capital poured into anything with

the word railway attached to it. And not

just from bankers, shopkeepers,

clergymen, widows, all of them piled

into railway stocks because nobody

wanted to be the one who missed out on

getting rich. At its peak, railway

investment hit around 7% of the entire

British economy. In a single year,

Parliament passed 272 separate acts just

to authorize new lines. Then came the

boom. Companies stopped building the

railways the country needed and started

building whatever they thought would

push their share price up. Routes got

duplicated. Lines got run out to

villages that would never fill a single

train. Then a reality check came when

the new lines were bringing in only

about a quarter of the revenue investors

had been promised. And once that was

undeniable, the math caught up all at

once. By 1850, the collapse was in full

effect and the British railway index saw

its shares being worth less than half

their peak. Fortunes that had felt

permanent were simply gone. And then the

last step, the inheritance. The original

investors got wiped out, but the rails

they had bankrupted themselves laying

were still sitting in the ground. Over

the following decades, that network

became the backbone of the entire

British industrial economy, and the

people who got rich were the ones who

scooped those assets up for pennies

after the bubble burst. That's the

capital cycle from start to finish.

Flood, boom, collapse, inheritance.

Money flooded toward the shiniest new

industry, certain that's where the best

returns were, until all the competition

led to oversupply, returns getting

crushed, and the people who supplied the

boom were destroyed by the very supply

they created. While the value they paid

for was left for whoever showed up after

the crash to snag for a great price. But

while the top hats and mutton chops make

it easy to dismiss all this as a

Victorian problem we would never fall

victim to now, that couldn't be farther

from the truth. Jump all the way to the

late 90s, and we can see the same cycle

play out again. After the US rewrote its

telecom laws in 1996, the flood poured

into fiber optic cable. Carriers buried

more than $500 billion of it, around 80

million miles, across the country and

under the oceans. But then the hype

started to fade, and reality caught up.

Around 85% of all that fiber sat in the

ground completely unused. Bandwidth

prices fell about 90%. The Nasdaq fell

more than 70%, and Corning, one of the

largest fiber makers in the world,

watched its stock go from over $100 a

share to about a dollar. And funny

enough, Corning over the past year has

become one of the hottest names in AI

right now, and their stock chart looks

awfully similar today as it did back in

the early 2000s. If we fast forward

another decade, we see the pattern play

out again in oil. Innovations in

fracking unlocked massive, previously

inaccessible reserves. Capital flooded

in. Drillers came chasing all at once.

US oil production roughly doubled, and

the US became the largest oil producer

on Earth. And then the supply blew

straight past the demand. Between 2014

and 2016, the price of oil collapsed

from over $100 a barrel to about 30.

More than 200 North American oil and gas

companies went bankrupt. The borrowers

drowned and the cheap energy they left

behind got picked up at a steep

discount. So, railway mania in the

1840s, fiber optic cable in the late

90s, and shale just over a decade ago.

Three different centuries, three

different technologies, three completely

different industries. Nothing in common

except one thing, the exact same cycle

playing out underneath all of it. Which

brings us to the biggest one yet, the

one our entire economy is currently

clinging on to for dear life, artificial

intelligence. This is the one that

matters the most because when the other

three collapsed, they took down some

investors and a couple of industries.

But artificial intelligence right now is

holding up the entire economy. Before we

get into just how big this AI thing has

actually become, a quick pause. Because

a lot of what I've covered so far comes

down to one pattern. The people building

the supply are the ones taking the risk,

the factories, the infrastructure, the

capital. They spend first and hope

demand shows up later. Which brings us

to the sponsor of today's video,

Printful. They do the opposite. Printful

is print on demand. You make a design,

put it on a product, and nothing gets

made until someone's already bought it.

Meaning no upfront inventory costs, no

stacks of boxes taking over your garage,

and no risk of being stuck with products

nobody wants. The order comes first.

Printful handles everything after that.

Here's what it looks like. I just take

whatever design I choose, upload it to a

product inside Printful, like this

coffee mug, connect to a store like

Shopify or Etsy, publish it, set up

billing, and that's it. The whole

process takes less than an hour and I'm

ready to start selling. With 508

customizable products to choose from,

more than a million items fulfilled each

month, and a 99% plus order approval

rate. Printful is the fastest, most

reliable way to turn an idea into a

product you can sell online. If you're

interested, you can check out Printful

using the link below or in the pinned

comment. Thank you to Printful for

sponsoring this video. And now back to

how the entire United States economy has

quietly become dependent on one massive

bet. A Harvard economist named Jason

Furman ran the numbers, and in the first

half of 2025, he found that investment

in information processing equipment and

software, which is the AI buildout, was

only about 4% of GDP, but it accounted

for roughly 92% of all the growth in the

US economy. So, if you strip the AI

buildout from the equation, the US

economy only grew at about 1/10 of 1%.

So, where does AI sit inside the capital

cycle? Well, the money is still pouring

in, but if you look closely, there are

cracks starting to show. Remember when I

said you can fake demand and that we'd

come back to it? This is us coming back

to it. In the '90s, they fake demand

with a story. WorldCom stood up and said

internet traffic was doubling every 100

days, and the whole industry ran with a

number that was never real. This time,

they're not faking it with a story,

they're faking it with money. You've

probably seen this chart that looks like

a circuit board. Every name on it is one

of the biggest companies in AI, and

every arrow is money moving between

them. A purchase here, an investment

there, a loan over there. And if you

follow almost any arrow, it loops back

to one place. Nvidia funds OpenAI.

OpenAI pays Oracle. Oracle buys Nvidia

chips. The money makes its rounds

throughout the industry and eventually

comes home to Nvidia, and every time it

passes a hand, it gets counted again as

somebody's brand new demand. So, a lot

of what looks like a booming market is

really just the same money going around

in a circle. If this sounds familiar, it

should. The fiber bubble ran the same

trick. Equipment makers like Lucent and

Nortel lent their own customers the

money to buy their gear, then booked it

as revenue. Activist hedge fund and

alternative asset manager Buxton

Helmsley clocked this immediately. They

noted that vendor financing did not

merely add risk at the margin. It

manufactured the appearance of end

demand. Equipment sold to customers who

could pay only because the seller had

financed them was recognized as revenue,

validated growth narratives, and

supported valuations until the moment

the sellers could no longer extend

credit. So, that circle can do a lot. It

can manufacture demand, prop up revenue,

and keep the chips moving. But, there's

one thing it can't do. It can't escape

the ideas Adam Smith laid out about

money chasing high returns, competition

entering, and returns getting compressed

for everyone. And the best place to look

for returns slipping is in the eye of

the storm, OpenAI. Two years ago, OpenAI

basically was AI. ChatGPT had no real

competition. Then everyone showed up,

Google, Anthropic, Meta, xAI, a dozen

others, all selling more or less the

same thing. It's the boom arriving,

right on schedule. Then last year,

OpenAI's financials leaked. For the

first time, we could see what that

competition did to the hottest company

in the world, and it's ugly. OpenAI lost

38 and 1/2 billion dollars in 2025,

nearly eight times its loss the year

before. But, the real story for OpenAI

isn't just in the headline number. It's

buried a few lines down. When you look

at what OpenAI had to spend on sales and

marketing, it jumped more than 400% in a

single year to 5.7 billion dollars.

That's 44% of their entire revenue spent

just to keep people using the product.

That's the cost of the boom. The returns

are falling, exactly the way they fell

for railways, for fiber, and for shale,

which means you already know what comes

next. And we're already seeing early

signals of it throughout the industry.

In early June, a chip maker called

Broadcom beat its earnings, but guided

its AI sales for the next quarter a few

hundred million dollars light. For a

company worth nearly 2 trillion dollars,

that should barely move the needle. But

following that news, about 300 billion

dollars of its market value vanished in

a single session. One of the biggest

single day losses in Wall Street history

over what amounts to a rounding error

for a company of Broadcom's size. That's

the tell. When a few hundred million in

soft guidance can erase 300 billion

dollars in a day, you're not looking at

a healthy market. You're looking at one

priced for perfection, where being even

slightly less than perfect costs you

more than a quarter of a trillion

dollars. There's an old line in

business. The pioneers get the arrows

and the settlers get the land. It means

the first companies into a new market

take the hits, the risk, the losses, and

the resistance. Then the later arrivals

walk in after everything has stabilized,

the dust has settled, and quietly take

the prize without ever taking the

arrows. You've watched this happen in

front of your eyes for your whole life,

whether you realize it or not. Take

search. In the mid-90s, AltaVista,

Lycos, and Excite were the pioneers.

They spent millions building the first

web crawlers and teaching the world what

a search engine even was. Then they

buried their pages in news, email, and

banner ads chasing revenue to pay for it

all and let the actual search product

decay. Then Google walked in, didn't

repeat their mistakes, and took the

land. Or ride share. Before Uber and

Lyft, a company called Sidecar spent its

life getting sued by cities and blocked

by taxi commissions fighting the legal

battles that made the model possible.

And while it was stuck in court bleeding

cash, Uber and Lyft copied the model,

raised billions, and took over the world

on the foundation Sidecar had built.

There's countless other examples of

this. It's the capital cycle theory. The

first movers flood in, prove the market,

and pay for it with everything they

have, then eventually get taken down by

the arrows. The supply they built gets

left behind for the settlers, where the

second movers walk in and inherit it for

pennies on the dollar. Which brings us

all the way back to where we started.

Everyone is staring at one side of the

coin, the demand side, screaming about

whether it's real. And it's easy to see

why. The demand side is loud, it's a

story, and a story is what gets clicked.

Traditional financial media will keep

you locked on that side for exactly that

reason. But in finance, the side that

actually decides what happens is never

the loud one. It's the quiet mechanics

underneath that drive everything. If you

want to track that side instead of just

following the noise in traditional

media, hit subscribe. Once you

understand the cycle of how capital

flows through industries, the whole

fight everyone is having online looks

like the wrong fight. Bubble or not was

never the right question. The question

was always who floods in, who drowns,

and who inherits. Once you start asking

the real questions, you start getting to

the real answers.

>> [music]